Tuesday, July 07, 2009

New centerboard (part 9) - Pouring in the real centerboard

With the test pouring done, we started making plans for the real thing.

In the test, as the lead cooled, it shrank and pulled away from the sides. To fix that, I wanted to use a smaller hole and slightly overpour, and then use a hammer to smush the middle, driving it towards the edges. Also, I read I should taper the edges of the hole, so the lead would flow around it, making the lead hold on to the shape of the wood. This would also smooth the transition when the lead shrank. In both cases, I needed a smaller hole, so I reduced from 6" square to 5" square for the first cut of the hole, with the plan to cut some more away for the taper.

The 1" spade drill bit provided the corners again. This time I used the jigsaw to cut the sides of the hole, since the circular saw wouldn't fit in the smaller space. I then turned the jigsaw to cut a 45 degree angle, and shaped the edges a bit. I followed with a rasp to take the edges closer to a 30 degree angle.

I didn't want to use screws on the inside edge again, since the lead hadn't flowed around them well. Nails seemed like a better choice. The plans show 3 nails per side, but getting them in without splitting the plywood seemed tricky, so I drove one in to each corner. Because the corners weren't tapered, they seemed stronger and needed help holding the lead in.

Because I wanted to overpour the lead for later pounding, I added a top mold, with a hole cut in the middle. The top mold also means that being perfectly level isn't required. So now the bottom layer is a 10" square; the middle layer is the centerboard with a 5.5" hole cut out, and the top layer is a 10" square with a 4" hole. A 1" pour hole would have been easier to cut (with my 1" spade bit), but harder to aim for. It would also have held in more heat, and I wanted the lead to cool faster so it would burn less wood.

I was told that lead was thin, so make sure everything is clamped tightly in place. I didn't have a clamp deep enough for the inside corner of the molds, so I used a lead-acid battery (in a case to protect from heat) to apply pressure in one corner. I've seen a section of railroad track used for this purpose. I'm not sure where to find one. Clamps did the other parts.

I decided to switch back to the cast iron pan instead of the coffee can. The coffee can didn't have a good pour spout, which was a problem during the test pour. I saw a lot of extra lead left behind in the can, so I used pliers to pull out some, and then melted out the rest. It took a long time to heat the can through the cast iron pan (pictured), so I removed the pan and heated the can directly. At this time I noticed the pan handle was so hot that I couldn't hold it long, even with the gloves. I timed it at ~10 seconds: not enough for a pour. I figured I could use the vise grips on the handle.

Once the extra lead was out of the coffee can, I put the cast iron pan back on, and then put the block of lead in. After 1/2 an hour of burning there wasn't any molten lead. It was breezy, which carried the heat away quickly. The cast iron pan had a lot of surface area and thermal mass, making it harder for the small campstove to do its job. Finally, the large block of lead was only touching the pan base at two small points, because the lead block was too big to fit in the pan. I needed to cut it like before, but now it was very hot.

I attacked it with a drill bit this time, drilling a series of holes in a line on one side, and then the other, and then poked at it until the lead split easily under my pliers. I gathered up all the shavings and put them in the pan, along with the two large blocks of lead. The shavings melted first, and helped transfer heat in to the block, which worked well.

To keep the heat from blowing away, I surrounded the setup with the panels from a plywood play house. When that didn't help enough, I added a piece of ceramic tile (from my "scary sharp" setup) on top of the pan, as a loose lid. Even better would have been heat reflectors, perhaps tile with foil over them.

After another round of waiting, the lead was melted. There was a little dross again, so I started skimming. After 5 minutes of "final" preparation, the lead had started to solidify at the edge of the pan. To work I had removed the plywood panel wind breaks and ceramic tile cover, and the wind was carrying away my heat.

Put them back in place, and waited again. Decided to do the pour quickly, so things would still be hot.

After another 15 minutes or so, the lead was all melted again. I clamped on the vise grips but found that it was too heavy to lift that way and stay under control. I used a second set of pliers to get both hands working together.

After the lead cooled, we removed the top mold, and saw that there was way, way too much lead. I shoulda stopped pouring when it was just ~1/16" up in to the top mold. Instead we had 1/4" or more. I experimented with removing it with a wood chisel and a block plane, but ended up buying a cold chisel to cut it away.

We spent a couple hours with hammer and cold chisel, to remove extra lead. Then we used the hammer to smooth the surface of the lead on the top, which caused it to bulge on the bottom. Hammered the bottom back to shape, then flipped and attacked the top again. After a while we decided it looked good enough. There was a gouge where we dug too deep at one point, and we were able to mostly repair it but not completely. Epoxy will have to fill in the hole.

I'm not certain if the centerboard will sink like it's supposed to. The new hole is considerably smaller than the old one (and the plans), and a substantial amount of lead is left over. However, the new centerboard is made of a denser, heavier wood, so it may be OK. If not, I can drill a few smallish holes and cast in the lead scraps.

Here's the good (bottom) side, in its final form.

- Clean up the cuts that went off the line
- Draw and cut the curve at the top
- Plane down the leading and trailing edges
- Cut a hole and pour in a lead sink weight (first time pouring lead!)
- Cut a gap for the pivot
- Prepare 1/8" sheet brass as pivot hole reinforcement (in progress)
- Drill 5 holes for fid and lanyard (ooh, the easy part)
- Epoxy and paint
- Install centerboard & lanyard
- Put boat back on trailer

Monday, July 06, 2009

New centerboard (part 8) - Failure becomes success

After failing to melt lead with a propane torch, I went looking for stronger tools.

Many instructions say to use a campstove to melt the lead. You want to be outdoors so the fumes don't collect (where do they go?). I didn't want to buy a new campstove for just one use, so I searched for used ones. There aren't many around. I finally found one at Marine Exchange, a used boating goods store. The salty old guy there suggested I use a weed burner, instead. It would produce a lot more heat, which meant I could melt the lead faster, using much less fuel. It took him 10 minutes of searching to find all the bits I would need: a hose with a regulator, a fuel switch, and fittings to adapt between each. Most of that time was finding the right combination of fittings. $10 for the set.

He also suggested a coffee can as a crucible, and vice grips to pick it up and pour. When I felt how heavy the cast iron pan + 11 lbs of lead was, I decided to get a coffee can. Who buys coffee in a can any more? I think most people buy it in plastic or fresh-ground in paper. It took a while to find one, but eventually I tried freecycle, and someone delivered it to my door. (Thank-you!). I didn't have vice-grips, so I decided to use 2 large pliers.

I borrowed a propane tank from a friend, and enlisted the help of my father-in-law who was visiting.

The block of lead was too big to fit in the mouth of the coffee can. My father-in-law said that a hacksaw would gum up easily with lead, and suggested a different approach. We used a hammer and a prybar, and perforated down the center. Flipped it over and repeated. Then leaned it against a concrete step and wailed with a sledge hammer, at which point the lead bent and we could pull it apart and put it in the coffee can.

We got all the propane hose bits hooked up, with Teflon tape between the fittings, and soap water to look for leaks. There were a lot of parts! Here's the setup, ready to go:

We fired up the weed burner and pointed it in to the can. The breeze blew the flame away from the can, so we made no progress fast. We moved in to the garage, with doors open for ventilation, and tried again. After 10 minutes we saw lead at the top melt, drip down, and freeze at the bottom. It looked like it would take a long time. I was getting tired of standing there holding the weed burner, so we turned it off and looked for a new plan.

Filled the propane tank up, to make sure there'd be plenty of fuel. (6.9 gal for $17). Went to the local general store for a camp stove, and found a basic model for $22. It was listed as 10,000 BTU, although I suspect those numbers are as much marketing as anything else.

I saw that it would screw in to the propane bottle that came with my little torch, but I was worried that there wouldn't be enough fuel for our job, and wanted to use the big tank instead. We looked through the set of fittings they had and picked one that seemed like it would work.

The next day we hooked up the camp stove to the big propane tank, went to light it, and got nothing. No fuel was coming out of the stove. We noticed that the small propane bottle looked like it had a pin on it that could open something on the stove's fitting, and figured we needed something more complex to push it. At the hardware store we found a different fitting that didn't just adapt to the right thread of for the camp stove, but included the little pin that we thought was missing. The hardware store also had a hose with the left-threaded fitting on one end, for a big propane tank, and a "throwaway" fitting on the other end for the camp stove. It was exactly what we needed, but at $42 I passed.

Back home we put the new fitting on, hooked it up, and again got nothing. *sigh*. After further pondering, we figured that there were two regulators involved, and that might be the problem. One built in to the hose, sitting right on the propane tank. The other in the camp stove's tubing. I guessed that they interfered with each other, and that stopped it from working.

We removed the regulator on the hose, and saw that it had different threads on each end. That meant I needed another adapter. I also saw that hose end was left-threaded, and there was a small adapter between it and the regulator.

We loaded all the equipment in to the van and drove back to Marine Exchange to find the right fitting. After 10 minutes we found two fittings that, together, would do the job. Drove back home, reassembled, and got leaks. Lots of leaks. We eventually figured out that one of the new fittings was really meant for plumbing, not for propane, and wouldn't seal well with the other fittings. Also, two of the fittings that should have worked together were leaking a little.

Here's the full collection of fittings:

We went back to the hardware store, and looked for alternative fittings. After 10 minutes of pondering, I decided to spring for the $42 hose that was exactly what I needed, instead of a cobbled-together collection of many parts that kept failing me in different ways. I wanted to get the lead melted already! I also picked up a pair of vice-grips, because having them securely clamped to the can seemed safer.

Put the lead in the coffee can on the stove, and turned it on. After about 1/2 an hour it was all melted. I used a cheap metal spoon from the thrift store to skim off the junk on the top. Poured it in the mold, saw it bubble and settle, and then let it set.

After it cooled I took it apart, and examined the result. The lead had pulled away from the edges all around, up to 1/8". It had not flowed over the tops of the screw heads; they probably cooled the lead quickly and were too high and thick. The bottom was relatively smooth and shiny, while the top was rough. During the bubbling stage, some tiny splashes of lead had left little bits sitting around, which we collected for later use.

New centerboard (part 7) - failed test pour

The centerboard is made of plywood, which is bouyant in water. We want it to stay down, so we have to add some weight. One way to do this is to melt and pour lead in to the board. The original was done that way, and I decided to do it again, reusing the lead.

Casting lead was all new to me, and it took a while to get things together. I decided to do a test pour first, and then repeat in the real centerboard.

To create the test setup, I started with two 10" x 10" pieces of the same plywood as the centerboard is made of. Neither was scrap, I'm afraid. The plans call for a 6" x 6" hole, to hold 10.9 lbs of lead. I drew a 6" square, then measured 1/2" in from each side. A small pilot hole, followed by a 1" spade drill bit in each corner got things started. To avoid tearing up the exit hole, I drilled 1/2-way through all 4 holes, then flipped and drilled the other side to match. Finally, used my small circular saw to cut the edges of the hole. That isn't very safe, as the saw wants to jump when the blade hits the surface. A jigsaw is safer, but straight cuts are harder and my circular saw is a better tool than my jigsaw.

Lead doesn't stick to wood, so I put nails and screws in the inside edge for it to mold around. I used both I wanted to see which would work better.

For a crucible (to hold the lead), I bought a 6" cast iron pan from the thrift store. Everyone recommends a coffee can, but I thought the built-in handle and pour spout would be useful.

I had a propane torch from a yard sale years ago, and decided to give it a go for melting the lead. It's a classic, probably from J. C. Penny. The instructions are amusing. In those days, you didn't need to worry about fumes from melting paint.

Applied the flame to the lead and all that happened was the lead and paint on the surface melted. I used the opportunity to scrape a lot away, but no lead melted. Not enough heat.

Sunday, July 05, 2009

New centerboard (part 6) - pivot plates.

The plans call for a pair of brass plates to reinforce the pivot hook on the centerboard. After popping them out of the old centerboard, I used a mill file to bevel the top edges, to smooth the transition to the wood.

The plans say to use screws to hold the plates in place, but the original didn't have them. Spent some time with the drill to machine some counter-sunk holes. Drilling brass it tricky, but I got the job done eventually. I have an ancient J. C. Penny drill press that didn't help much. I gave it away a few days ago, and will get a good one if I decide I really need one.

Used a mirror-sharp chisel to cut slots to bed the plates in the centerboard. If the plate had been placed on the surface of the plywood, all shear forces would be taken by the screws. The plates are 1/8" thick. The 3/4" plywood is a little thinner than advertised, and cutting it down 1/8" on each side would leave less than 1/2" of wood to take the strain, in an area where strength matters. So I chiseled through 3 plies, which makes them about 1/16" deep. Seems like a good compromise. The plates fit very snugly, which helps make sure they can transfer the shear forces to the wood.

I also decided to use some Gorilla Glue to hold the plates to the board. It is strong stuff, and grows by 4x as it cures, and sticks to just about everything.

I practiced screwing in to scrap plywood and found that it was easy to drive too hard and strip out the wood in the hole. I decided to use my bit and brace ($3 at a used building supplies store!) to do the last turns of the screws. I also experimented with different size pilot holes to see what works the best.

So, the installation procedure was:
  1. Wet both the wood and the metal plates
  2. Dab on a very thin layer of Gorilla Glue
  3. Wait 5 minutes for things to get tacky (gold lamé?)
  4. Install one plate
  5. Drive screws with the drill most of the way in
  6. Use the brace to finish the screws off
  7. Flip and install the other side.
It was at "flip" that we saw a problem: the screws were slightly too long. The screw tips poked through the wood on the other side, which would stop the other plate from fitting properly. Meanwhile, the glue is drying...

I quickly grabbed a Dremel grinder bit (I don't have a Dremel, just a grinder bit, left over from making a lock pick set 12 years ago), tossed it in the drill, and ground down the tips of the screws. They are soft brass, so it went fast. Then I put the other plate in place and clamped it down while the glue cured.

Meanwhile I cut the tips off the other screws so they wouldn't be too long. I used the grinder to shape them to a little point, to help the screw get started. Gorilla Glue reaches 80% strength in a an hour or two. After lunch I unclamped and drove in the remaining screws.

The result worked out OK, although I do see one corner bulging because of a screw that's still too long. Oh, well, it's too late to change and will be OK.
From Jay Bazuzi's personal blog
As you can see, the slots in the brass overhang the wood by about 1/16", so they take the strain first.

- Clean up the cuts that went off the line
- Draw and cut the curve at the top
- Plane down the leading and trailing edges
- Cut a hole and pour in a lead sink weight (first time pouring lead!)
- Cut a gap for the pivot
- Prepare 1/8" sheet brass as pivot hole reinforcement (in progress)
- Drill 5 holes for fid and lanyard (ooh, the easy part)
- Epoxy and paint
- Install centerboard & lanyard
- Put boat back on trailer
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.